Universal Access To All Knowledge
Home Donate | Store | Blog | FAQ | Jobs | Volunteer Positions | Contact | Bios | Forums | Projects | Terms, Privacy, & Copyright
Search: Advanced Search
Anonymous User (login or join us)
Upload

Reply to this post | Go Back
View Post [edit]

Poster: JFLL Date: May 17, 2009 6:37am
Forum: texts Subject: Searching for Periodicals

The Internet archive has a wonderful collection of old periodicals, but sometimes they seem to want to make it as hard as possible to find things. For example, there is a journal called 'Archaeologia Cambrensis' of which 93 issues are on IA. I wanted to look at the issue for 1882, but for every single issue the title on the first page is just 'Archaeologia Cambrensis' - no hint of a volume number or date! What's more, they seem to be in random order. So I had to go into 93 individual records, and then I couldn't find the 1882 issue. (It seems they've uploaded 1883 twice, and missed out 1882.) Please, uploaders, think about your users!

This post was modified by JFLL on 2009-05-17 13:37:11

Reply to this post
Reply [edit]

Poster: stbalbach Date: May 17, 2009 8:55am
Forum: texts Subject: Re: Searching for Periodicals

JFLL, I agree with you in general. In this case, you could have searched and found the book with this search string:

title:"Archaeologia Cambrensis" year:1883

If you look at the metadata for any of the books in this series (click on HTTP then go to the file "name_meta.xml) you'll see there is a field for the year, which you can then use to build the search. But not all books have this "year:" field, just lucky in this case.

This sort of underscores the nature of Internet Archive. Books come in from all over the places, various libraries, Google - the metadata is inconsistent, you have to learn how to do complex searches and navigate its peculiarities.

There are two solutions:

1. Internet Archive pays someone to clean up the meta data and make it consistent. I'd rather IA spend its limited money scanning more books personally.

2. We, the end users, make our own meta data. I did this recently for the `Notes and Queries` periodical which can be seen here:

http://docs.google.com/Doc?id=dc2wxqtd_4d3gzwrdb

Reply to this post
Reply [edit]

Poster: garthus Date: May 17, 2009 8:28pm
Forum: texts Subject: Re: Searching for Periodicals

Stbabach,

I agree with you. The prime directive is to preserve as much information as is possible under the circumstances. Much of the stuff which I am putting up may well disappear since it is next to impossible to find in public libraries or archives. It stick as much of the ID stuff in the key words field as I can. Once the stuff is up, the meta data can always be repaired at a later date. Of course it would be nice to have perfect meta data, but this is to much to expect of volunteers and in many cases even of professionals in the field. We should all be thankful for the many hours of work by volunteers building this collection (The Internet Archive); I for one find it much more useful than the commercial data bases out there and think that it can only get better with each passing day.

Gerry

Reply to this post
Reply [edit]

Poster: stbalbach Date: May 18, 2009 9:28am
Forum: texts Subject: Re: Searching for Periodicals

Gerry and JFLL,

For periodicals in general, John Mark Ockerbloom over at onlinebookspage.com has done a lot of work indexing. He is "site agnostic" (ie. not just Internet Archive) so has links from all over.

http://onlinebooks.library.upenn.edu/serials.html

He recommends that interested persons would "adopt" a serial/periodical, which includes indexing it (among other things), then use his site as a central place to post the list in a regular format - I've just sent him my `Notes and Queries` list since his is incomplete. It's a slightly awkward system as you have to submit a form to do updates (and it appears backlogged), and much of it is out of date. A Wiki that anyone can edit would be a better way to do it IMO.

Stephen

Reply to this post
Reply [edit]

Poster: MaureenKennedy Date: Jun 23, 2009 11:37pm
Forum: texts Subject: Re: Searching for Periodicals

The above link has it's own inherent flaws - I wanted to look up Harper's Magazine - so it's list send me to Cornell Univeristy where I have to log in, second link sends me to Harper's Magazine where for $17 I can look up old records from their mag. Huh? So much for free access.

Reply to this post
Reply [edit]

Poster: JFLL Date: May 18, 2009 1:34am
Forum: texts Subject: Re: Searching for Periodicals

Thanks very much, Stbalbach and Garthus, for your help and comments. In fact I found after posting that the 'sort results by date' filter seemed to work in this case. Of course I agree with both of you that the main aim must be to get as much material on IA as possible, but the year and issue of a journal are fairly crucial, and should always be tagged so as to come up on the search results page, surely. Mine was really a plea for a consistent, and helpful, treatment of journal entries.

Reply to this post
Reply [edit]

Poster: garthus Date: May 18, 2009 6:19am
Forum: texts Subject: Re: Searching for Periodicals

JFLL,

You are 100% correct, the poster merely has to place the year in the date field, for example, one of my postings:

http://www.archive.org/details/The_Primary_Class_11_2_13

This item has a volume, number and article; identification. By sticking this info in the title and keyword search your search would easily bring up the item with only partial information.

Gerry

Reply to this post
Reply [edit]

Poster: MaureenKennedy Date: May 24, 2009 8:06am
Forum: texts Subject: Re: Searching for Periodicals

If you opened 93 issues, then you could have written 93 "reviews" saving other people the effort you went to, by describing what issue was the one you are looking at.

Reply to this post
Reply [edit]

Poster: JFLL Date: May 24, 2009 9:31am
Forum: texts Subject: Re: Searching for Periodicals

Well yes, I could, but then I'd have wasted even more of my time! My point was that users shouldn't be put to this inconvenience in the first place -- it's the job of the uploader to tag titles helpfully, by thinking about the effects of his/her tagging for likely search strategies. Librarians know (or knew) all about this, which is why they devised cataloguing rules a hundred or more years ago.

Reply to this post
Reply [edit]

Poster: stbalbach Date: May 24, 2009 4:41pm
Forum: texts Subject: Re: Searching for Periodicals

Two things:

1. These books in particular came from Google. Google seems to have limited metadata info available for the public to access. This may be intentional.. but not sure.

2. Internet Archive has some sort of automated bot that is copying on the order of 500,000 to 1 million books from Google to Internet Archive. So, that's the "uploader/librarian" in this case.

So I don't think this is a case of a lazy individual or bad practices (on IA's behalf). Why isn't Google more forthcoming or accurate with its metadata is the question. My guess is Google as a commercial company makes money from metadata (searching), they don't want to give away the store.. but not sure.

Anyway, all this comes back to a commercial vs non-profit approach .. open vs closed .. This is just one example of the type of issues that come up when your dealing with a corporation that has all sorts of opaque motives and rules and regs. Is a darling of Wall Street the best Librarian? Probably not.

Stbalbach

Reply to this post
Reply [edit]

Poster: JFLL Date: May 25, 2009 12:40am
Forum: texts Subject: Re: Searching for Periodicals

I'm sure you're right, Stbalbach. Good tagging/cataloguing needs trained staff, and trained staff = money. I'm reminded of a certain music download subscription service I use, where tagging is often chaotic, with the result that all sorts of weird search strategies have to be used to find what they have by a given composer. They obviously don't see adequate and consistent tagging as being important to their "business model". Very irritating for an ex-librarian!